14 Pediatric Nuclear Medicine: A Coming of Age Helen R. Nadel

14 Pediatric Nuclear Medicine: A Coming of Age

Helen R. Nadel and Moira E. Stilwell

H. R. Nadel, MD, FRCPC;

M. E. Stilwell, MD, FRCPC

Department of Radiology, University of British Columbia, Children’s and Woman’s Health Center of British Columbia, 4480 Oak Street, Vancouver, B.C. V6H 3V4, Canada

A discussion of pediatric scintigraphic techniques would not be complete without mention of technical factors required due to the unique needs of imag- ing pediatric patients who range in age from birth to adulthood. Advances in radiopharmaceutical and instrumentation technology allow greater investiga- tion of physiology as well as anatomy in this diverse group of patients.

14.1.1

Patient Preparation

It is routine in many departments to allow parents and/or siblings to remain in the imaging room to provide a sense of security and safety for the child.

The presence of a favorite toy or a prized possession brought with them for the test can also be reassur- ing. Immobilization needs, which can be an impor- tant factor in producing high-quality studies in children, can be broken down by age. For neonates to age 2 years, bundling in a papoose-type holder, sleep deprivation, and feeding the child while on the imaging table are effective immobilization strat- egies. Videos, music, stories, or distraction tech- niques, such as bubble blowing, thus ensuring their cooperation, entertain children aged 4–5 years. It is also helpful to have dedicated pediatric nuclear medicine technologists who are secure in their deal- ings with the various age groups. A simple explana- tion of the involved procedure tailored to the age of the child will go a long way to gain a child’s coopera- tion; however, some children still require sedation or general anesthesia. In particular, those children who are mentally retarded or have severe attention deficit problems, and often children between the ages of 2 and 4 years, fall into this category. If seda- tion is necessary, appropriate triage, monitoring, and discharge of the patient are now the standard of care (Weiss 1993; Pintelon et al. 1994). The use of a local anesthetic cream at an intravenous access site is also routine in many laboratories (Sherazi and Gordon 1996; Ljung 1997).

CONTENTS

14.1 General 227

14.1.1 Patient Preparation 227

14.1.2 Radiation Safety Considerations 228 14.2 Helicobacter pylori Infection in Children 228 14.2.1 Diagnostic Tests 230

14.3 Imaging of Inflammatory Bowel Disease 232 14.4 Pediatric Neuronuclear Medicine 234 14.4.1 Scintigraphy in Brain Development 235 14.4.2 Language and Cognition 237

14.4.3 Brain Trauma and Brain Death 237 14.5 Primary Neuro-Psychological Disorders

in Childhood 239

14.5.1 Attention Deficit Hyperactivity Disorder 239 14.5.2 Mood Disorders: Depression

and Obsessive–Compulsive Disorder 240 14.5.3 Tourette Syndrome 240

14.5.4 Autism 240 References 241

14.1 General

Nuclear medicine studies contribute an integral

part to the investigation of many disease processes

in children. The standard studies of bone, renal,

and tumor imaging with scintigraphy are still the

backbone of any pediatric nuclear medicine depart-

ment. Over the past decade we have seen the intro-

duction and increased use of newer scintigraphic

approaches in children. This chapter focuses on the

use of scintigraphy in children for the assessment

of Helicobacter pylori infection and inflammatory

bowel disease and psychosocial, developmental, and

traumatic brain disorders in children.

14.1.2

Radiation Safety Considerations

The use of ionizing radiation in the pediatric popu- lation is an ongoing issue frequently cited in the pediatric radiology literature and bears examination by all practitioners. Roebuck (1999) examines and re-examines a variety of related issues. He points out rightly that part of the job of the radiologist and/or nuclear medicine physician is to help ensure that the appropriate investigation is performed, and that use of radiation is justifiable. Knowledge of the approximate risk, and more importantly the ability to effectively communicate these to the referring physician and parent, is essential. Nuclear physi- cians and radiologists alike should be aware of new applications of nuclear medicine tests but remain skeptical (Roebuck 1999; Berdon 1999; Hall 1999). Discussions regarding the risks of radiation and what exactly they consist of are frequently mis- understood and misquoted even by the physicians using them. In a letter to the Lancet, Dixon and Dendy (1998) posed three valuable questions to be considered before performing a CT scan on a child.

These questions can be rephrased for nuclear medi- cine tests: Can the same or better information be obtained by a method with a lower dose of radiation?

Is the utilized technology functioning properly? Is the protocol optimized for the patient? Stabin and Gelfand (1998) have published tables for evaluating absorbed radiation doses for many pediatric nuclear medicine procedures. They used standard MIRD methodology and express absorbed dosimetry in mGy or rad for a body weight adjusted administered activity for each radiopharmaceutical.

14.2

Helicobacter pylori Infection in Children

The battle cry of pediatric medicine is that “children are not just small adults who suffer from small ver- sions of adult diseases.” Whereas this is true mainly in pediatric gastrointestinal disease, many entities that we commonly see in adults are seen in children and need to be considered when appropriate. One such entity is Helicobacter pylori (H. pylori) infec- tion. H. pylori infection is endemic worldwide as an asymptomatic infection (Narla et al. 1999). The recent discovery and growing interest in this gram- negative bacterium as a gut pathogen has had and continues to have significant impact on the man-

agement of several gastrointestinal disorders. Peptic ulcer disease and gastritis have been closely linked to H. pylori infection, and this infection is impli- cated in a substantial percentage of cases of acute gastrointestinal bleeding. H. pylori has been clas- sified as a type-I carcinogen by the World Health Organization (Narla et al. 1999). In addition, it has been impugned as a co-factor in other gastroin- testinal diseases such as chronic diarrhea, protein- losing enteropathy, and gastric lymphoproliferative disease. Corrado et al. (1998) studied 90 patients and found a positive association between H. pylori infection and food allergy.

H. pylori appears to be acquired early in life with cross-sectional studies suggesting that infection is usually acquired before the age of 5 years, although the time, rate, and modus of colonization are not definitely defined (Bujanover et al. 1996). Mater- nal passive transfer of immunity via IgG antibod- ies appears to protect infants until approximately 6 months of age when infection of children with H.

pylori is common in both developed and developing countries (Gold et al. 1997). There is correlation with socioeconomic and hygiene conditions such as overcrowding and lack of clean running water (Duggan et al. 1998; Vandenplas and Blecker 1998). The role of parental infection is not fully understood but is thought to play a role in transmis- sion particularly in infected mothers of young chil- dren (Rothenbacher et al. 1999). H. pylori infec- tion in children is thought to be affected by both the rate of acquisition and loss in different groups based on ethnicity. A study by Malaty et al. (1999) followed 212 children in a biracial community over 12 years from childhood to young adulthood and showed different rates for both acquisition and loss of infection in groups of black children compared with groups of white children. To illustrate, at 7–

9 years of age, 19% of children in their population had H. pylori infection (40% of black children and 11% of white children). Twelve years later, 22% of the children were positive with the higher preva- lence again noted to be present in the black chil- dren. Granström et al. (1997) studied a cohort of Swedish children from ages 6 months to 11 years of age measuring both IgG and IgA antibodies to H. pylori. They identified that 13.6% (40 of 294) of their children were infected at some time during the study. They noted that at 11 years of age only 3%

of the children were seropositive suggesting that

spontaneous resolution is common. In a study done

in Germany, Hornemann et al. (1997) could not

find evidence of H. pylori infection in infants less

than 4 years of age. After age 4 years, the seroposi- tivity increased linearly with age.

H. pylori infection is a well-recognized cause of antral gastritis and a significant factor in peptic ulcer disease in the adult population. In children, peptic ulcer disease does occur, albeit less commonly (Stringer 1989; Hassall 1994; Huang et al. 1999).

Peptic ulcer disease in children is mainly duodenal ulcer disease, and when H. pylori is the cause, antral gastritis is also present. H. pylori can cause isolated gastritis as well. When H. pylori induces antibod- ies and reacts with them in the gastric mucosa, chronic gastritis may result and it is thought that these mucosal changes play a role in the progres- sion to gastric cancer in adults. Ieradi et al. (1998) found that autoreaction of gastric mucosa may be found in H. pylori gastritis of childhood. They pos- tulated a role in progression to atrophy and possi- bly increased risk of late gastric cancer. Numerous properties of the bacteria itself are also suggested as being important, including the presence of the CagA protein and vacA alleles associated with cyto- toxin production. H. pylori generally falls into one of two groups: a Cag-negative group which does not produce cytotoxin and a Cag-positive which does and is associated with peptic ulcer disease (Çelik et al. 1998). Çelik et al. (1998) confirmed that the Cag pathogenicity and toxin expression seen in adults is also seen in younger subjects with recur- rent abdominal pain. Children and adolescents demonstrate 40% CagA positivity as compared with 80–100% of adults with duodenal ulcer and gastri- tis. Mitchell et al. (1999) also found that the pres- ence of CagA is not a marker of specific disease in children. Furthermore, these investigators conclude that children do become persistently infected by H.

pylori and few carry strains with the virulence prop- erties seen in adults with gastrointestinal pathology.

Whether these children should be treated to reduce risk of gastritis and gastric cancer is not clear at this time (Çelik et al. 1998). The current view holds that H. pylori is spread to children from close person-to- person contact and, because there is familial clus- tering seen with infection of children, it has been suggested that family treatment may decrease treat- ment failures due to increased compliance (Fall et al. 1997; Oderda et al. 1997).

H. pylori is responsible for approximately 80%

of duodenal ulcer after other etiologies including Crohn’s disease (CD), non-steroidal anti-inflam- matory drug ingestion, and Zollinger-Ellison syn- drome have been ruled out. In a retrospective review of pediatric gastroscopy carried out in Hong Kong,

88% of patients complaining of dyspepsia had posi- tive histological findings. In this group 25% were positive for H. pylori (Ng et al. 1997). Heaney et al.

(1998) have suggested that H. pylori infection status can be used as a criterion to select patients for gas- troscopy when determining management of young dyspeptic patients. In this study 160 “young” patients (<45 years of age) were prospectively recruited when presenting with the complaint of epigastric pain and/or discomfort. Patients who were negative for H. pylori using the carbon-13 (

C) urea breath test were reassured that they would likely have a normal gastroscopy. At this point they were given lifestyle advice and symptomatic therapy and then followed at intervals of 6 weeks, 3 months, and 6 months. If they were the same or worse at these intervals, they proceeded to gastroscopy. This management proto- col resulted in a 42% reduction in the number of gas- troscopy examinations in H. pylori patients.

Recurrent abdominal pain is a common present- ing complaint in children. There is no substantial evidence to suggest that H. pylori has any role in recurrent abdominal pain or so-called non-ulcer dyspepsia in children. Wewer et al. (1998) assessed the IgG seroprevalence of H. pylori in children with recurrent abdominal pain compared with healthy children to investigate the presence or absence of related symptoms. IgG antibodies against low molecular weight H. pylori were assessed in a total of 438 children and 91 healthy controls. The sero- prevalence in children with recurrent abdominal pain was 21%, against 10% in the controls with 46 of 66 of those positive by culture and histology. The presence of H. pylori was associated with both par- ents being born in a country with high prevalence of H. pylori, low socioeconomic class, and more pain related to meals but not related to low levels of hemoglobin, leukocytes, thrombocytes, weight, or height. No specific symptomatology was related to H. pylori positivity in children. In a case control study of Turkish children by Günel et al. (1998) no association was found between children with H.

pylori infection, hypergastrinemia, and recurrent abdominal pain. H. pylori infection was as high in healthy children as in those with recurrent abdomi- nal pain confirming the findings by Gottrand et al. (1997) that normal gastric histology can be asso- ciated with infection in children.

The wide interest in this bacterium has spawned

many investigations of its relationships, if any, with

other gastrointestinal diseases, with inflammatory

bowel disease (IBD) being a case in point. The cause

or causes of IBD remain incompletely understood

with infectious agents including several bacteria and viruses implicated. Since H. pylori infection causes changes in both permeability and immunological arrangements in the stomach similar to those seen in the colon in IBD, it is natural to wonder if there is a causal relationship. Currently, there is no epidemi- ological data to support that conclusion and while, interestingly, there is a low prevalence of H. pylori in inflammatory bowel disease, the clinical signifi- cance, if any, of H. pylori infection against the back- ground of IBD is not yet clear (Nardone et al. 1998;

Parente et al. 1997). Permanent colonization of the stomach in pediatric patients with IBD is unusual (Kolho et al. 1998). In a cross-sectional study with matching of patients with IBD and upper gastroin- testinal lesions it is noted that the reduced H. pylori infection prevalence is due mainly to decreased frequency of colonization in patients with CD. The prior treatment of patients with sulfasalazine but not 5-aminosalicylic acid, steroids, or immune suppres- sants is associated with a reduced risk of infection in both patients with ulcerative colitis (UC) and CD. In the group of patients with CD (123 of 216), 10% had gastroduodenal localization of their disease while 15% had H. pylori negative gastritis on histological examination (Parente et al. 1997).

Celiac disease is a disorder frequently associ- ated with gastritis. In a study performed to assess the prevalence and significance of H. pylori infec- tion in children with celiac disease no increase in prevalence or clinical manifestations of H. pylori was found (Luzza et al. 1999). Recurrent abdomi- nal pain was the only symptom that distinguished between H. pylori positive and negative children, but this symptom resolved on gluten withdrawal regardless of their H. pylori status.

As H. pylori is a causal agent for gastritis, the presence of H. pylori infection in gastric-body- type mucosa as is seen in Meckel’s diverticulum may be pathogenic. Hill and Rode (1998) reported the case of a 25-year-old man who presented with rectal bleeding and a positive Meckel scan. Surgi- cal pathology revealed a diverticulum with gastric- body-type mucosa with active chronic gastritis and positive immunohistochemistry for H. pylori. They found only one other case in a retrospective review at their hospital of 21 cases of proven Meckel’s diver- ticulum.

In another line of investigation of gastritis, Sala- rdi et al. (1999) looked at the prevalence of H. pylori in children with type-I diabetes mellitus. Children with diabetes are more susceptible to common acquired bacterial infections and chronic atrophic

gastritis if they have longstanding disease. They con- cluded that, although children with diabetes do not differ from their healthy cohorts in terms of preva- lence of H. pylori infection in the first few years of disease, they subsequently do have a higher H. pylori seroprevalence. They suggest that this could be the cause or one of the causes of chronic atrophic gas- tritis in children with longstanding diabetes. When Barrio et al. (1997) looked at a group of children and adolescents with diabetes, they found that 7%

(12 of 177) of these patients had gastric parietal cell autoantibodies and 6 of 8 of those who had antral biopsies were found to have H. pylori infection lead- ing them to suggest that diabetic children and ado- lescents should be screened for infection.

In a study looking at the role of H. pylori in diminished growth in children, three populations of children were studied including normal, diabetic, and hyposomic children (Vaira et al. 1998). They looked at the possible role of infection by differing strains of H. pylori. As infection with more cytotoxic strains, such as the so-called CagA or VacA strains, lead to more severe gastroduodenal infection, these authors looked for a possible role for these strains in children with decreased growth, but they did not find supportive data.

14.2.1

Diagnostic Tests

When H. pylori infection is diagnosed in children and thought to be the cause of the patient’s morbid- ity, a course of double antibiotic therapy is recom- mended for 4–6 weeks’ duration (Hassall 1994).

H. pylori infection diagnosis can be achieved using invasive methods, such as endoscopy, to provide tissue samples for rapid urease tests, culture, and histological examination. Noninvasive testing can be performed by blood sampling for IgG antibody tests and urea breath testing using radioactive carbon-14 (

C) or stable

C. Clinical guidelines for the diagnosis and treatment of H. pylori in children currently suggest a fairly limited use in the young age group. Ongoing research into its role in gas- trointestinal disease and expanding use of nuclear medicine modalities in children may spur further utilization of these methodologies and knowledge of these for use in children will be valuable.

The noninvasive breath tests are painless and

can be performed in children who are old enough

to follow simple instruction for ingestion and blow-

ing through a straw. Both

C and

C urea breath

tests for the diagnosis of H. pylori infection can be utilized (Cutler 1997). Both tests take advantage of the fact that H. pylori produces urease which is needed to break down labeled urea. Gastric urease activity can then be measured and used as an indica- tor of the presence of infection. Both tests rely on the measurement of labeled CO

exhaled by the patient after consuming a standard dose of labeled urea.

Urea breath tests have the ability to sample a larger area of the gastric mucosa than individual biopsies.

In the

C breath test the patient is fed a high-carbo- hydrate, high-protein meal to slow gastric emptying and increase the residence time of

C-urea in the stomach.

C-urea is then ingested as a powder that is dissolved in water. Collection of breath samples occurs at baseline and at 30 min. Thirty minutes is the proposed collection time to avoid false-posi- tive results due to urease in the mouth flora. A mass spectrometer is used for the measurement of

C.

In the

C-urea breath test the elimination of

C via exhaled air is measured with a liquid scintilla- tion counting technique. The advantage of

C over

13

C is that smaller concentrations can be detected as radioactive substances are not routinely exhaled.

Baseline breath samples are not required; and there is no need to slow gastric emptying. After a 4-h fast, the pediatric patients are given 0.055 MBq dose

orally supplied as dissolvable capsules and samples of exhaled air are taken at as early as 10 min after oral administration with cumulative samples taken at 60 min. Caglar et al. (1999) have shown accurate results when comparing a single 10-min sample to 60-min cumulative excretion, thus improving cost- effectiveness and improving patient acceptance.

The accepted upper limit of normal for excretion at 10 min is a specific activity of a breath sample corrected for body surface area of 0.3% (Henze et al. 1990; Debongie et al. 1991). An abnormal test result would be a cumulative value of greater than 2% excretion.

In all methodologies the accuracy of the results can be reduced if certain drugs have been taken by the patient. Antimicrobials should be discontinued a minimum of 4 weeks prior to testing; bismuth- containing compounds 1 month prior; proton pump inhibitors and sucralfate or similar com- pounds at least 1 week prior to the test; histamine (H2) receptor antagonists for 2 days prior; and ant- acids for 1 day prior. Histological method accuracy is reported to be reduced in the presence of an active or recently bleeding duodenal ulcer (Cutler 1997).

Although there is some controversy as to which of the available methods of diagnosis of H. pylori infec- tion in the untreated patient is best, the urea breath

Fig. 14.1a,b. a Normal distribution of

Tc-HMPAO labeled white blood cells in a child. Anterior view of the abdomen and pelvis. Note the prominent activity in the spleen, liver and bone marrow. b infl ammatory bowel disease (IBD); dif- fuse involvement of the colon, including proximal ascending colon, transverse colon, splenic fl exure, and descending colon to sigmoid colon in a 10 year old girl with Crohn's disease.

Note the high uptake in the spleen in all three planes. MIP = maximum intensity projection image, i.e. a volume rendered summation of the maximal counts from the tomographic images, for the transverse, sagittal and coronal planes

a

b

ant post

1. Transverse MIP 2. Sagittal MIP 3. Coronal MIP

test is recognized as the method of choice to confirm H. pylori cure (Cutler 1997). Confirmation of erad- ication of H. pylori infection is most accurate when the breath test is performed a minimum of 4 weeks after the completion of therapy.

The

C-urea breath test is simple, easily per- formed, and inexpensive. It provides quantitative information regarding the total activity of the H.

pylori bacterium in the stomach as reflected by urease activity and severity of gastritis and is a sen- sitive and specific method of confirming its presence in peptic ulcer disease (Henze et al. 1990; Debongie et al. 1991). The test has shown between 95 and 98%

reproducibility when repeated after a 1-week inter- val in a series of children (Steen et al. 1995).

The

C isotope is radioactive and therefore raises considerations relevant to the use of ionizing radia- tion in the pediatric population . When

C-urea is used it is excreted in the breath as

CO

or in the urine. Biokinetic models have been established to determine dosimetry in the pediatric population (Leide-Svegborn et al. 1999). Approximately 88%

of the administered activity is excreted in the urine within 72 h of administration. Approximately 3% is exhaled within 20 days. The bladder wall receives the highest dose of radiation. In children aged 7–14 years this effective absorbed dose measures approximately 0.14–0.36 mGy/MBq. The effec- tive total body dose in children is calculated there- fore to be 0.9–2.5 µSv for an administered dose of 0.055 MBq. This is a minimal dose and therefore the use of this examination as a screening test and repeated examinations would be acceptable from a radiation burden standpoint.

14.3

Imaging of Inflammatory Bowel Disease

The demonstration of intra-abdominal sources of infection remains a challenging clinical problem.

The front line imaging techniques of ultrasound, CT, and MRI provide structural detail but cannot always differentiate adequately treated residual morpho- logical alterations. A variety of scintigraphic studies can provide physiological information not available through other means. This provides a significant contribution to the clinical problems of whether active infection exists in the abdomen, its location, extent, and response to therapy over time. Whereas gallium has a time-honored place in the visualiza- tion of infection, radiopharmaceutical innovations

allow more specific and more sensitive delineation of infectious processes.

The use of radiolabeled autologous white blood cells (WBC) to assess intra-abdominal sites of infec- tion is increasing. Labeled WBC will leave the cir- culation at sites of infection and migrate into the infected tissues as part of the normal host response.

99m

Tc-HMPAO as a practical label for autologous WBC offers high-quality images with favorable dosimetry.

In-WBC scanning is not recommended for children because of its unfavorable dosimetry (Gainey et al. 1988).

Tc-HMPAO-labeled WBC are proving to be particularly helpful in pediatrics for the assessment of inflammatory bowel disease (Charron 1997; Barabino et al. 1998).

Normal distribution of labeled WBC occurs in the spleen, liver, and red bone marrow (Fig. 14.1), with some faint activity in submandibular glands, kidneys, and bladder. Transient WBC sequestration in the lungs can occur in the first few hours. The gallbladder can be seen in up to 10% of children, but this does not interfere with interpretation. Some activity in normal ascending colon can be seen at 4 h post injection. This activity is usually faint and less intense than iliac crest activity and has a dif- fuse pattern. This is thought to represent excretion of non-cell bound labeled secondary hydrophilic complexes (Charron 1997). Since imaging for the assessment of bowel inflammation is not usually obtained beyond 4 h, this does not decrease speci- ficity.

No bowel preparation is required for this exami- nation, unlike with gallium imaging. The admin- istered dose is adjusted for body weight between a minimum–maximum dose of 185–740 MBq (5–20 mCi). Imaging protocol includes 0.5- to 1-h and 2- to 3-h static images in anterior and poste- rior projections and SPECT with low-energy high- resolution collimator. Tail on detector views can be used to help distinguish perianal and rectal disease from bladder activity. Upright anterior views of the abdomen can be helpful in separating liver activity from that in transverse colon. All imaging should be performed after the patient has voided. In children suspected of having CD, a view to include anterior chest and mouth can be used to assess for disease in the esophagus (Charron 1997). Charron et al.

(1997) visually grade the inflammatory activity in

each segment relative to the iliac crest bone marrow

and liver activity: grade 0=no activity; grade 1=less

than crest; grade 2=equal to crest; grade 3=greater

than crest; grade 4=less than liver; grade 5=equal

to liver; grade 6=greater than liver. Abnormal activ-

ity in keeping with inflammation appears bowel shaped, does not change configuration on the late scan, and increases in intensity temporally. The SPECT images reviewed in cine mode with maximal intensity can also increase sensitivity for detection of disease, although there may be some reduction in specificity.

Labeled WBC scintigraphy is particularly help- ful for the assessment of small bowel disease. The demonstration of disease distribution influences the selection of therapeutic regimens and has prog- nostic implications. Labeled-WBC scintigraphy also has the ability to detect extra-intestinal sites of inflammation, such as abscess collections and fistulae (Fig. 14.2). The labeled-WBC scan has also been able to discriminate between continuous and discontinuous disease thereby effectively differenti-

ating UC from CD (Charron et al. 1998). In a series of 106 patients with either UC or CD only 6 patients were classified incorrectly. Labeled-WBC scintigra- phy allows for accurate assessment of the segmental distribution of disease when correlated with his- tology and endoscopy (Charron et al. 1998). This technique can be particularly helpful in identifying sites of inflammation when technical difficulties preclude this assessment with colonoscopy. Labeled- WBC scintigraphy is less invasive than barium follow-through examination or enteroclysis and is much better tolerated by children. Limitations of labeled-WBC scintigraphy include the inability to assess anatomical detail such as bowel strictures and prestenotic dilations. Fistulae may be difficult to detect. If the abnormal activity is focal, it may be difficult to distinguish large from small bowel due

Fig. 14.2. 16 year old female with infl ammatory bowel dis- ease and previous partial colectomy with stoma in right iliac fossa and a fi stulous skin tract in the anterior lower abdominal wall. CT showing extensive thick walled bowel in right lower quadrant (c). Surgical abdominal scar in the midline marked with a paper clip anteriorly, where the clinically identifi ed skin fi stula was noted. The CT scan does not clearly iden- tify communication of the tract with bowel.

Tc -HMPAO labeled-WBC scan shows extensive abnormal activity in the right lower abdomen with visualization of the transverse colon. There is abnormal activity to the anterior abdominal wall seen on the planar views (a) and SPECT MIP images (b).

Fistulogram study showing contrast injected through a cath- eter placed through the skin fi stula fi lling a connection to the colon in the right iliac fossa (d). This appearance is consistent with a colocutaneous fi stula. MIP = maximum intensity projec- tion image, see Fig. 14-1b

a

c d

b 1. Transverse MIP 2. Sagittal MIP 3. Coronal MIP

to insufficient anatomical landmarks. In the child who is actively bleeding there may be difficulty in scan interpretation (Charron et al. 1999). Overall accuracy of 93% with 90% sensitivity and 97% spec- ificity has been reported (Charron et al. 1999; Del Rosario et al. 1999). Charron et al. (1999) report a positive predictive value for labeled-WBC scintigra- phy of 97%. In the evaluation of 79 control subjects no false-positive results were obtained. Negative predictive value was 93%. The disease distribution can be underestimated in early CD and nonspecific colitis. When there is focal bowel involvement in the region of the cecum, appendicitis as well as involve- ment of the cecum with IBD must be included in the differential. Although endoscopy and biopsy are still the gold standards, labeled-WBC scintigra- phy is helpful in the follow-up of IBD in detecting recurrence, monitor treatment, or plan surgery. This technique can reduce the need for repeated colonos- copies and barium examinations in these children.

14.4

Pediatric Neuronuclear Medicine

Just as there have been advances in the area of brain scintigraphy in adults, so too have there been advances in pediatric neuro-nuclear medi- cine. The advent of new radiopharmaceuticals for brain imaging has allowed scintigraphy to become a full-fledged member of the multidisciplinary team investigating neurological and psychological illness in the past decade. Cross-sectional imaging with CT and MRI revolutionized the treatment and diagno- sis of neurological disease in adults and children;

SPECT and PET imaging bring physiological and functional information to vastly improved morpho- logical imaging modalities. This has opened new doors for medical care providers and scientists to study and learn more about neurological disease, neuropsychiatric disorders, brain development, and learning in childhood.

Advances in instrumentation and radiopharma- ceuticals have made it easier to perform brain scin- tigraphy and to obtain more diagnostic informa- tion in children. The studies have focused mainly on regional cerebral blood flow, but new research is exploring metabolic and receptor imaging particu- larly in neuropsychiatric disease. The main indica- tions for pediatric neurological nuclear scintigra- phy include seizure disorders, assessment of brain development, evaluation of tumors, and trauma.

Knowledge is growing rapidly and expanding our understanding of developmental and neuro-psycho- logical disorders.

While heredity is a large contributor to brain structure and function, we know that development and learning after birth can and will have enormous impact on a child’s ultimate motor sensory and intel- lectual potential and capabilities. When birth or early developmental experiences have a potentially negative impact, parents look anxiously to physi- cians and allied health personnel for aid. Neuroim- aging can be performed to guide intervention, treat- ment, assessment, or prediction of ultimate outcome (Denays et al. 1990; Kerrigan et al. 1991; O’Tuama and Treves 1993; Kao et al. 1994; Yamada et al.

1995; Lee et al. 1998).

Radiopharmaceuticals used for PET have included

18

F-fluorodeoxyglucose (FDG),

C-L-methionine for tumor imaging, oxygen-

O for cerebral blood flow assessment of intraventricular hemorrhage and infarction, and hypoxic ischemic encephalopa- thy in infants. SPECT radiopharmaceuticals include

99m

Tc-HMPAO and

Tc-ECD or less commonly

123

I-iodoamphetamine or

Xe for regional cerebral blood flow assessment. Ligands labeled with

I or

11

C have been used in receptor imaging studies.

Tl has been used to differentiate recurrent disease from radiation necrosis in children with brain tumors.

Neuroimaging is also used in infants and children to learn about the normal and abnormal development of the brain. Neuroimaging is used more and more in the research and investigations of childhood psychiat- ric and behavioral disorders, which raises important questions about the value of these tests to children as a group and the individual and group attendant risks.

Investigations using ionizing radiation are justified

by using the calculation of the so-called risk/bene-

fit ratio of the test. This ratio may not be calculable

in the short term and certainly not for a particular

child at a particular time in development, assuring

that this debate will continue in the future. Ernst

(1999) points out correctly that the risk/benefit ratio

is essential to the conduct of research and requires

the calculation of risks according to the ambiguous

definition of “minimal risk.” When discussing the

risks of exposure to low-level radiation, it is helpful to

compare the risk to everyday activities that the child

might realistically expect to partake in, such as riding

in a car, as well as risks which might be uppermost in

the mind of children or their parents such as the risk

of developing leukemia or other cancers and the risk

of increased chromosomal abnormalities. In another

review article assessing health hazards to children

of radiation exposure in the context of brain imag- ing research, the authors conclude that health risks from low-level radiation obtained by radiographic and scintigraphic diagnostic studies are not detect- able above random events that occur in everyday life (Ernst et al. 1998).

Brain imaging research has huge potential bene- fits to society as a whole. Understanding its attendant risks to children is crucial to rational discussion of investigation of neurological and psychological dis- orders. Members of ethics and grant review boards, as well as individual researchers and practitioners, need to understand the risk of exposure to low-level radiation through research and diagnostic studies.

This research needs the participation of both chil- dren who are abnormal and those who are normal.

It is not unreasonable to assert that normal children will also benefit from further understanding of neu- rological and psychological dysfunction. PET stud- ies in normal child volunteers have to date recorded very low acceptable radiation exposures (Ernst et al. 1998).

14.4.1

Scintigraphy in Brain Development

Brain development has been extensively studied using PET and FDG. Similarly SPECT radiophar- maceuticals have been used to assess changes in regional cerebral blood flow with brain develop- ment in the normal human infant (Chugani et al.

1987; Chiron et al. 1992; Chugani 1998; Kinnala et al. 1996; Schiepers et al. 1997; Rubinstein et al.

1989). Brain development is phylogenetic with devel- opmentally older structures visualized before newer structures. At a gestational age less than 40 weeks, regional cerebral blood flow imaging shows promi- nent activity seen in the thalami with no parietal or occipital cortical activity and poor if any frontal activity. At 40 weeks or term gestational age, tha- lamic activity remains prominent. There is increase in activity in the parietal cortex, still low occipi- tal activity and poor frontal activity (Fig. 14.3a).

At 44 weeks of gestational age there is a promi- nent parietal cortex visualized, occipital activity is increasing, the thalami remain prominent, and there is still poor frontal activity. At 2 months of age there is predominance of parietal and occipital cortical activity with still low frontal activity. By 6 months of age there is marked cortical predomi- nance. There is increase in frontal cortex activity but still considerably less than the corresponding

parietal and occipital cortical activity (Fig. 14.3b).

By 1 year of age all cortical areas show a significant increase in regional cerebral perfusion with either PET or SPECT.

Cerebral glucose metabolism matures through childhood until typical adult activity patterns are seen at approximately age 1 year (Chugani et al.

1987; Chugani and Phelps 1986). Cortical glu- cose metabolic rate measured with PET shows an increase until age 2–3 years which can exceed adult values. This remains high until 8–10 years of age and declines to normal adult values somewhere between 16–18 years of age. This period of rapidly increasing glucose utilization corresponds to a time of overproduction of nerve terminals and synapses.

The plateau phase when glucose utilization remains stable but higher than that of adults represents over- connectivity followed by elimination of the exces- sive connectivity and corresponding reduction in cerebral glucose utilization to adult levels.

Takahashi et al. (1999) using PET have deter- mined normal values for regional cerebral blood flow (rCBF), regional cerebral metabolic rate for oxygen (rCRMO

), and regional oxygen extraction fraction

Fig. 14.3.

Tc-HMPAO activity pattern in a normal new- born (a) and at 6 months of age (b). Note the initial lack of frontal activity, prominent brain stem and temporo-parietal perfusion, which increases over the fi rst six months with more uniform cortical activity seen at 6 months

a b

(rOEF). The children in their series ranged in age from 10 days to 16 years. In the neonatal period the rCBF and rCRMO

were lowest and increased during early childhood. A temporal sequence of rCBF and rCRMO

similar to that seen with glucose metabo- lism and brain SPECT findings was noted, likely reflecting physiological development within ana- tomical areas of the brain.

Chiron et al. (1992) studied 42 neurologically normal children of all ages (2 days to 19 years) with normal CT, EEG at time of study, and normal psychomotor development for 2 years post study using rCBF by SPECT with

Xe and compared the results to an adult reference group consisting of 32 subjects 19–29 years of age who were free of neuro- pathology. The mean rCBF was calculated for three age groups and the measurements showed effects of age in all cortical regions and on global blood flow. Global blood flow averaged 50 ml/min per 100 g at birth and increased in all age groups to a maximum at 5 years and decreased to adult levels after 19 years of age. Cortical regions all demon- strated lower perfusion at birth than in adulthood, and all regions peaked at around 5 years of age at levels 70% higher than adult levels. When studied on a relative basis, rCBF was at a lower level at birth than at maturity and increased in the first 2 years of life to plateau during the rest of childhood with less than 10% variation. In addition to refining reference values for large functional areas of the brain, this work reconfirmed the importance of the early years of life in neurological development.

Brain function assessed by regional glucose uti- lization and PET in the neonate is highest in the thalamic nuclei, brain stem, cerebellar vermis, and sensorimotor cortex which correlates with the domi- nance of subcortical activity at that age. Early physi- cal examination of an infant elicits the basic reflexes such as rooting and the grasp reflex. As development progresses, the regional utilization of glucose con- tinues to correlate with the development of function in corresponding anatomical regions and these are observed in PET studies in the parietal, temporal, and primary visual cortex, basal ganglia, and cer- ebellar hemispheres described as encephalization during 2–4 months of age (Chugani 1992).

While work in this area is ongoing, other efforts are aimed at using the neuroimaging techniques of SPECT and PET to attempt to study and, hopefully, predict outcome of neurological insult and catas- trophe in the very young. One group of children in whom this would be helpful is the pre-term infant.

These infants often have protracted courses in hospi-

tals with superimposed serious illnesses which may conclude in death. Those who survive have often been subject to long periods of artificial ventila- tion with varying levels of low p

O

and high p

CO

, low hematocrit, and low arterial blood pressure, all of which may negatively impact on development and growth. It has been reported that ventilation in neonates is associated with low CBF and spe- cifically with hemorrhage and neurological deficit (Greisen 1986, 1990; Greisen and Pryds 1988;

Allan and Volpe 1986; Altman et al. 1993). With respect to intracranial hemorrhage, cranial ultra- sound is an effective investigation in the preterm infant for diagnosis and for predicting outcome.

Hemorrhage tends to be less damaging to the very young infant than infarction into the periventricu- lar white matter or periventricular leukomalacia.

Infants born prior to 30–32 weeks gestational age are at risk due to the thin-walled vessels in those areas and combined factors of decreased cerebro- vascular reserve and regulation. Baenziger et al.

(1999) studied brain perfusion in these infants in an attempt to elucidate the predictive capacity of CBF measurements in this group. A total of 71 pre- term infants (less than 1500 g birth weight and less than 34 weeks gestational age) were studied at three time intervals postnatally: between 2 and 36 h; between 36 and 108 h; and between 108 and 240 h respectively using

Xe and extracranial cadmium telluride detectors. Corresponding cra- nial ultrasound was performed. The surviving chil- dren were assessed for development at 18 months of age. There was a higher mean CBF and mental or motor development. Other studies in children with cerebral palsy have a spectrum of findings on brain SPECT studies which include absence of cerebral blood flow abnormality to thalamic hypoperfusion, diffuse hemispheric hypoperfusion, focal hypoper- fusion in the contralateral hemisphere to an area of motor deficit, and cerebellar areas of hypoperfusion (Denays et al. 1990; Kao et al. 1994; Lee et al. 1998).

Pryds et al. (1990) suggest that cerebral hyperper- fusion in term asphyxiated infants who lose their ability to autoregulate may be an earlier indicator of cerebral hypoperfusion. There is no definitive pattern identified in children with birth asphyxia which can predict long-term sequelae, but cerebral blood flow imaging studies are helpful in the identi- fication of CBF disturbances.

The brain and skull can be involved in a variety of syndromes or isolated insults in early infancy. Cra- niosynostosis is one such developmental anomaly.

When there is premature fusion of multiple sutures

or involvement of only one suture, the result is brain compression which can lead to impaired cognitive development. Surgical correction can include cra- nial vault remodeling in severe cases and or strip craniectomy. Pre- and post-operative FDG PET studies in a small number of affected children have shown varying regional increases and decreases;

however, more consistent postoperative increase in activity has been demonstrated in the posterior occiput in the region where visual development and visual–spatial coordination develops (David et al.

1999).

14.4.2

Language and Cognition

One of the main and best established uses of func- tional brain imaging is the localization of epilepto- genic foci in children in whom surgical treatment is planned. When CT and MRI have ruled out a morphological abnormality, radionuclide perfu- sion imaging inter- and postictally offers a sensitive method for localization of the seizure activity. The finding of decreased regional perfusion at a focal area and increased perfusion at the same site during a seizure offers a sensitivity of approximately 95%, although this number falls if temporal lobe epilepsy is excluded (Gordon 1996). Nuclear physicians may be unaware that there are substantial considerations for these children regarding language problems and behavioral abnormalities which must be understood to allow the best possible treatment.

There are many language and behavioral manifes- tations associated with epilepsy and not infrequently these may be epileptic events which are erroneously ascribed to behavioral problems (Tuchman 1994).

Although the anatomical basis linking seizure dis- orders to disorders of behavior, such as aggression, depression, and even schizophrenia, have been pos- tulated, this is not yet universally accepted. These areas are fertile fields for functional imaging. The total effects of seizures on the brain are incompletely understood and still debated (Devinsky and Bear 1984; Holmes 1991; Lesser et al. 1986). In the 1950s Landau and Kleffner (1957) described a syndrome of acquired aphasia in epilepsy which has been fol- lowed by numerous other similar reports, although the exact pathogenesis and neuroanatomical basis for this is unclear. In a study designed to demon- strate the regional cerebral perfusion in Landau- Kleffner syndrome and its correlates with the EEG, O’Tuama et al. (1992) performed brain SPECT using

99m

Tc-HMPAO in five children with this syndrome and in three with other speech difficulties. They found perfusion asymmetry with decreased perfu- sion in the temporal lobes most marked in the peri- sylvian cortex. A PET-FDG study performed in three boys with this syndrome showed similar findings, mainly in the temporal lobe distribution. Neither study nor subsequent studies have yet defined the correspondence to EEG abnormalities (Maquet et al. 1990).

Congenital dysphasia has also been studied using SPECT. This is a clinical diagnosis of exclusion where deafness, mental retardation, speech pathol- ogy, and neurological and psychiatric causes have been ruled out. In a study of 14 affected children, SPECT demonstrated decreased activity in the left hemisphere, best localized near Broca’s area in 2 patients with expressive aphasia. In 9 of 12 children with both expressive and comprehension problems there were two areas of abnormality best localized to the left temporo-parietal area and the right frontal lobe (Denays et al. 1989).

Children suffering from epilepsy have a higher risk of learning and behavioral disorders, which is multi-factorial and can be very complicated to inves- tigate. Many of the important factors are not directly relevant to functional brain imaging; two of interest to nuclear medicine specialists are seizure activity and the location of the epileptogenic focus both of which have significant negative effects on cognition.

Specifically, temporal lobe seizures are considered a definite risk factor for cognitive developmental problems and functional brain imaging studies have confirmed this through a suspected negative effect on verbal memory (Aldenkamp et al. 1990).

14.4.3

Brain Trauma and Brain Death

The determination of brain death relies on unequiv- ocal clinical data indicative of irreversible cessa- tion of brain function, specifically both cortical and brain stem functions. Supportive laboratory studies, most importantly two silent EEG studies separated by various time periods depending on legal juris- diction, are desirable and often required. In cases of catastrophic injury of a child, the declaration of death can be a highly charged situation for the relatives and physicians involved in the child’s care and for physicians involved in harvesting and trans- planting organs into other ill children and adults.

Not infrequently, an objective test provides comfort

for this task. Younger et al. (1989) claim that knowl- edge of medical and legal criteria for diagnosing death among physicians is surprisingly lacking. It is important that the contribution of nuclear brain scans be available, prompt, and knowledgeable.

Cerebral perfusion studies using tracers, such as

99m

Tc-HMPAO or

Tc-ECD, are a quick and accurate alternative to traditional gold standard confirma- tory four-vessel angiogram examination, to confirm absence of intracranial perfusion and help provide a timely diagnosis of brain death. These radiopharma- ceuticals have advantages over the blood-brain bar- rier agents formerly used in that there is less difficulty in differentiating intra- and extracerebral activity and activity in the dural sinuses is not a potentially con- fusing picture. Technical aspects of bolus injection and timing of the study are less stringent and assess- ment of the posterior fossa using the

Tc-lipophilic perfusion agents is far easier (Fig. 14.4; Laurin et al.

1989; Schiepers et al. 1997).

The absence of perfusion above the tentorium with preservation or persistence of blood flow in the cerebellum and/or brainstem contradicts the diag- nosis of brain death, although it has a grave prog- nosis (Valle et al. 1993). The planar scintigraphy should include a lateral view and SPECT should be performed for the best evaluation of the posterior fossa. A prudent recommendation is to include views of the thyroid bed to confirm radiopharmaceutical quality. Views of the liver and heart also support

radiopharmaceutical integrity and an estimation of viability of those organs (Wieler et al. 1993). Imag- ing can also be done with a portable camera in the ICU setting, if desirable. All patients showing absent brain uptake with these tracers are confirmed as brain dead uniformly in published series. Reported series include mainly adult patients (Larar and Nagel 1992; Wilson et al. 1993; Goodman et al.

1985). The diagnosis of brain death in children can be difficult and adult criteria are not necessarily sufficient or easy to obtain (Holzman et al. 1983).

There are several series of pediatric patients which confirm that absence of intracerebral perfusion is a highly reliable aid in diagnosing of brain death (Fig. 14.4). This is particularly useful when the clini- cal criterion of silent EEG is complicated (Galaske et al. 1988).

Nonfatal head injury ranging from mild to severe has also been studied using lipophilic

Tc-labeled brain agents and FDG. Both CT and MRI scans as the front line tools in the diagnosis of head injury can be less sensitive than nuclear medicine studies. Trau- matic brain injury abnormalities in regional cerebral blood flow and glucose metabolism can be docu- mented, more extensive, and detected earlier than with CT or MRI (Abu-Judeh et al. 1998). Evidence of ischemic injury is common and can be present in any region of the brain. It is attributed to diffuse or focal axonal shearing in the white matter in the absence of a structural abnormality on CT or MRI scan.

Fig. 14.4. Brain death. Posterior (a), left lateral (b) and anterior (c) views obtained with planar scintigraphy. There is absence of normal intracranial activity on these

Tc-HMPAO images. d Right lateral view of a brain scan showing normal intracranial perfusion after administration of

Tc- HMPAO

b

d

a c

Regional hypoperfusion on SPECT can be concor- dant or discordant with reduced glucose metabolism on PET (Abu-Judeh et al. 1998; Yamaki et al. 1996).

When the findings are discordant, it indicates a good prognosis which may be due to vasospasm and edema causing decreased perfusion which is temporary.

Most studies demonstrate abnormalities in the tem- poral frontal or parietal lobes and correlate well with severity of injury generally described by duration, severity of coma, and length of retrograde amnesia (Jacobs et al. 1994; Gordon 1996). An initial negative SPECT examination within 4 weeks of a closed head injury is a robust predictor of full recovery, although a positive scan is of less use prognostically. The role of SPECT and PET brain studies in the clinical and medico-legal work-up and assessment of head injury and its sequelae has not been fully established. Efforts to gain a toe-hold are hampered by lack of knowledge and understanding of the usefulness and applicability by medical practitioners during the acute and chronic phases of head trauma evaluation (van der Kolk 1997). Furthermore, whereas early studies focusing on acute head trauma included patients in the pedi- atric age group, there are few studies exclusively of children. Studies focusing on the chronic sequelae of brain injury when relationship between the morpho- logical findings may be difficult to confidently relate to the neurological, psychological, or behavioral find- ings, are also sparse. Goshen et al. (1996) published a series of 28 pediatric patients with chronic sequelae of traumatic brain injury. These patients were referred to the rehabilitation unit of the medical center. They ranged in age from 15 months to 16 years and were assessed using EEG,

Tc-HMPAO scintigraphy, CT, and, when available, MRI following substantial head injury (Glasgow coma scale ratings of 3–10).

Retrospective analysis showed that cerebral blood flow studies following brain trauma were more sensi- tive than the anatomical modalities for detection of abnormalities, particularly in the basal ganglia and cerebellum. These authors suggest that brain SPECT may play a role in children who have neurological sequelae in the follow-up of brain trauma.

14.5

Primary Neuro-Psychological Disorders in Childhood

Using the classification of O’Tuama et al. (1999) there are four primary categories of childhood neuro-psy- chiatric disorders: attention-deficit hyperactivity

disorder (ADHD); mood disorders such as depres- sion and obsessive–compulsive disorder (OCD);

Tourette syndrome and related disorders; and infan- tile autism and autistic spectrum disorder. One of the most inspiring uses of functional brain imaging must be the study of neuro-psychological disorders in children. The definition, etiology, diagnosis, and treatment of many of these disorders can be difficult and controversial. Few would dispute the negative outcome on adult life nor deny the often alarming repercussions that many of these disorders have on emotional development and family relationships.

The ability to study cerebral perfusion and metab- olism with radionuclides has inspired great hopes for better understanding of psychiatric and devel- opmental disorders in children and adolescents. The hope has been to aid or substantiate the correlation of neuro-psychological deficits or behavioral abnor- malities with specific morphological regions and metabolic abnormalities. However, as pointed out by Filipek (1999), behavioral problems in children provide a spectrum of disorders which encompass deficits or derangements in language, cognition, visual–spatial function, and socialization.

14.5.1

Attention Deficit Hyperactivity Disorder

Attention deficit hyperactivity disorder continues to receive enormous attention in the lay media and is a subject of great ongoing study and research. The diagnosis has evolved substantially in recent years in tandem with advances in genetic, cognitive, and pharmacological research among other factors. The American Psychiatric Association includes the fol- lowing criteria as assessed by reports by parents and teachers: poor impulse control; hyperactivity; poor temper control; and poor attention span. This dis- order is relatively common, affecting up to 5–6% of children, and its propensity for boys is well known.

The syndrome manifests in adulthood and leads to substantial social problems and alienation. Affected children are generally further classified as predomi- nantly inattentive, predominantly hyperactive, and a combination type. Research using neuroimaging is confounded by frequent overlay of other behav- ioral disorders, mood, and anxiety disorders, and learning disabilities. Although the exact mechanism for this is not certain, advances in cognitive research in this area consider the main impairment in ADHD to be the inability to inhibit a behavioral response.

Multiple models exist in the literature pointing to an

arising focus of cognitive and genetic research, and neuroimaging (Tannock 1998).

Neuroimaging, specifically MRI, suggests that developmental abnormality involves the right hemi- sphere and basal ganglia (Tannock 1998; O’Tuama et al. 1999). Brain SPECT imaging neither confirms nor refutes global changes in cerebral perfusion and currently most convincingly presents hypoperfu- sion and, therefore, presumably decreased neuronal functioning in the striatal regions (Lou et al. 1989).

More recent work using PET has demonstrated decreased glucose metabolism in the pre-frontal cortex (Tannock 1998). Experimental lesions in animals in both the striatal and prefrontal regions may cause hyperactivity (Lou et al. 1989). Although the available studies can be described as embryonic, at best, there is a clear indication that it is appropri- ate for nuclear brain imaging to join the research in this important disorder.

14.5.2

Mood Disorders: Depression and Obsessive–Compulsive Disorder

One of the most important advances in childhood neuropsychiatric disorders is simply the acknowl- edgement and understanding that mood and anxiety disorders do occur in young children and should be treated seriously when they exist. Functional brain imaging has shown great promise in the investiga- tion of mood disorders. Decreased perfusion and metabolism has been demonstrated by SPECT and PET in the left prefrontal cortex and limbic areas, most notably in recurrent or chronic depression. In contrast, patients with transient or normal sadness demonstrate increased activity in the same areas when they are asked to recollect sad events (Risch 1997). This supports the theory that this activated circuitry in normals persists when it becomes chronic and ultimately burns out. Functional imag- ing in depression in adults most frequently demon- strates decreased perfusion in the temporal, frontal, and parietal areas. There is a significant shortage of studies in children in whom depression is now well recognized. Applying adult criteria to children may be completely erroneous as changes in brain activ- ity may fluctuate during the course of an affective illness as well as during normal brain development (Risch 1997; O’Tauma et al. 1999).

The same caution applies to OCD which occurs in adults and children but has mainly been studied in adults. This disorder is not uncommon in chil-

dren and teenagers and can be debilitating by virtue of intrusive and disturbing obsessive ideation and compulsive behaviors. As in other neuropsychiat- ric illnesses, a biologic or biochemical basis is sus- pected, although it is not known where this occurs or what its exact nature is. Functional brain imaging studies suggest involvement of the basal ganglia as well as the orbital–frontal cortex (Swedo et al. 1989;

O’Tuama et al. 1999).

14.5.3

Tourette Syndrome

Tourette syndrome is a neuropsychiatric disorder that presents with motor and vocal tics most promi- nently along with frequent associations with ADHD, obsessive-compulsive disorder, and other disturb- ing behavioral symptoms. Both SPECT and FDG PET studies have demonstrated hypoperfusion in frontal and temporal lobes and basal ganglia regions (Lampreave et al. 1998; Sieg et al. 1993). As in the other conditions discussed herein, the ages of the patients can confound results with changes due to maturity and/or evolution of the disease with or without co-morbid features. Data from older age groups should not be assumed to apply to children in general, but they certainly suggest that theory- based studies applied to children may be helpful.

As in other disorders the observed hypoperfusion is assumed to be linked to decreased neuronal activity.

Improvement observed with neuroleptic treatment may be due to decreased hyperactivity in the dopa- minergic system leading to clinical improvement (Lampreave et al. 1998).

14.5.4 Autism

Autism currently referred to as autistic spectrum or pervasive developmental disorders is a wide-rang- ing gamut of deficits involving verbal and non- verbal behaviors. Pervasive developmental disor- ders includes autistic disorder, Asperger syndrome, and Rett syndrome among others (O’Tuama et al.

1999; Filipek 1999).

Rett’s syndrome is a well-known but rare neuro-

degenerative disorder which affects very young chil-

dren, and is only seen in girls. In an effort to better

understand the biological basis for this tragic dis-

order, PET studies have been performed using FDG

and D2 receptors agonists, although conclusions are

hampered by the small number of subjects. Naidu et al. (1992) demonstrated decreased activity in the occipital cortex in association with slightly increased activity in the frontal cortex. Yoshikawa et al. (1991) studied six patients and suggest that impaired oxi- dative metabolism exists in Rett syndrome. Using

15

O

PET they found that the cerebral metabolic rate of oxygen was decreased in five patients and oxygen extraction fraction in four when compared with three normal controls. Both measures declined with advancing age. The loss of hyperfrontality is also stressed in a study with 13 patients (Lappalainen et al. 1997). SPECT-ECD brain imaging confirms decreased perfusion in the frontal and fronto-pari- etal cortical areas (Burroni et al. 1997). These per- fusion abnormalities become more marked in the later stages of the disease. In general, the perfusion abnormalities precede abnormal findings on MRI.

Gross brain anatomy in subjects with autism is generally normal and neuroimaging studies are inconsistent or inconclusive. The relationship of cor- tical migration abnormalities to this spectrum of dis- orders, if any, is not yet known. MRI studies in high functioning adults with autism, known as Asperger syndrome, have demonstrated abnormalities includ- ing polymicrogyria, macrogyria, and schizencephaly without a predilection for a particular region in the brain (Filipek 1999; Chakos et al. 1998). Findings from neuropathological and neuroimaging studies in groups of patients with autism can be generally sum- marized as: increased brain volume with patients demonstrating increased head circumference when compared both with normals and patients with other developmental abnormalities; impaired function at the corpus callosum; and impaired frontal lobe func- tion and abnormalities in the areas of the mind where socialization responses and understanding occurs (Filipek 1999; Deb and Thompson 1998). Data from activation studies would be of potential value in investigating social responses. Data from SPECT and PET studies is only available for a small number of patients with autism many of whom are adults and are further hampered by lack of technical uniformity.

Numerous findings are generally not reproduced in other studies (Müller et al. 1999)

Infantile autism is currently described as a developmental disorder for which no specific neu- robiological or pathological correlate is defined.

Schifter et al. (1994) evaluated retrospectively 13 children with the diagnosis of autism having dys- functional social interaction skills, decreased or abnormal verbal abilities, and abnormally limited interests and abilities. Unlike other studies, they

did not exclude those with significant other abnor- malities such as seizure disorders and mental retar- dation. These children ranging in age from 4.5 to 11 years were studied using FDG PET and CT or MRI, and no common or uniform abnormality was found in the group. The most frequent abnormality found was decreased FDG uptake and for the most part this existed in patients with seizure disorders.

The authors feel justified in excluding interictal foci as a cause of this finding due to the multiplicity and multifocal nature of the abnormalities. Further- more, they conclude that the findings support other work impugning neuronal migration abnormalities, at least in part, as a significant feature. This study illustrates the myriad of technical and logistical dif- ficulties in these and related types of disorders. The sensitivity and specificity of the various imaging studies, although interesting, are far from estab- lished but should not discourage ongoing efforts of nuclear physicians to better understand this disease and contribute potential solutions to it.

Functional neuroimaging in childhood disorders continues to be challenging and is complicated by the fact that only pooled data are available, which provides valuable albeit frequently inconsistent information. It is important to understand the large variety of issues to ensure that studies are performed in a manner most likely to enhance and complement ongoing advances in cognitive research, genetics, and therapies. Defining subject groups is difficult and made worse by frequent overlay of more than one disorder. Control groups are often not available in children and subject groups to date are very small. When radionuclide studies are performed, there is variation in scanning techniques, methods of analysis, and activation protocols. None of these factors should discourage informed investiga- tors from further pursuing functional brain imaging research in children.

References

Abu-Judeh HH, Singh M, Masdeu JC, Abdel-Dayem HM (1998) Discordance between FDG uptake and technetium-99m HMPAO brain perfusion in acute traumatic brain injury. J Nucl Med 39:1357–1359

Aldenkamp AP, Alpherts WCJ, Dekker MJA, Overweg J (1990) Neuropsychological aspects of learning disabilities in epi- lepsy. Epilepsia 31:S9–S20

Allan WC, Volpe JJ (1986) Periventricular-intraventricular hemorrhage. Pediatr Clin North Am 36:47–63

Altman DI, Perlman JM, Volpe JJ, Powers WJ (1993) Cerebral

oxygen metabolism in newborns. Pediatrics 92:99–104

Baenziger O, Mueller AM, Morales CG, Jaggi JL, Duc G, von